Description of Data and Code for Maurer, Klein and Vella.

The data file mkv_data_final.fmt (GAUSS format) contains the final data underlying the analysis in Maurer, Klein and Vella. The data set has 10,854 rows (1,809 individuals with 6 observations each) and 16 columns (one column for each variable). Additional information on the data can also be obtained from the original data source, the RAND HRS Data, Version D.

The variables are ordered as follows:

1.	Individual identifier 
2.	Currently working for pay (dummy)
3.	Age (in years)
4.	Presence of any mobility limitation (dummy)
5.	Presence of any large muscle activity limitation (dummy)
6.	Presence of any ADL limitation (dummy)
7.	Depression: CESD score>4 (dummy)
8.	Presence of a major health conditions (dummy)
9.	Presence of a minor health condition (dummy)
10.	Married or partnered (dummy)
11.	Partner working (dummy)
12.	Partners� self-rated health (ordinal, 5 point scale)
13.	Manual occupation (dummy)
14.	Years of education
15.	Race: Black (dummy)
16.	Race: Hispanic (dummy)

The main GAUSS codes are Estimation1.g for the estimation of coefficients and their variance and ASFvarest.g for the estimation of the ASF and its variance. The other code provided are procedures called by these programs. 